Software Engineer - ML Ops

Exol•Wilmington, MA

23h•$96,000 - $132,000•Hybrid

About The Position

With its A.I.-powered robotic technology platform, Symbotic is changing the way consumer goods move through the supply chain. Intelligent software orchestrates advanced robots in a high-density, end-to-end system – reinventing warehouse automation for increased efficiency, speed and flexibility. The ML Operations Team at Symbotic is expanding, and we’re looking to hire a software engineer to help us build computer vision solutions. The candidate will assist in the development of computer vision models that will help our fleet of autonomous robots to accomplish their missions safely, accurately, and with super-human efficiency. The engineer is expected to help in all tasks that would develop computer vision solutions and deploy them to run on our fleet.

Requirements

Bachelor’s degree or above in computer science, software engineering, robotics engineering or related field
Demonstrated experience with cloud-based infrastructure and services, including Google Cloud Platform (GCP) or Microsoft Azure
Experience working with complex data systems, including data storage, telemetry data, and camera or sensor data
Strong software engineering skills with expertise in at least one programming language and its ecosystem. You should be comfortable delivering high-quality, working code quickly and consistently.
Proficiency in Python or C++.

Responsibilities

Architect and scale our software platform to support current and future autonomous capabilities.
Collaborate with ML R&D teams to ensure new models meet deployment requirements.
Optimize the robotic hardware fleet deployment pipeline, streamlining processes from merge requests through testing to roll-out for systems using ONNX/TensorRT and custom hardware.
Maintain and improve edge computing infrastructure to meet real-time performance and reliability needs.
Create pipelines to train robotic machine learning systems including dataset curation, labeling, training, and evaluation metrics.
Double speed and efficiency of robotic system evaluation using simulation of hardware and mechanics
Reduce mechanical hardware maintenance during robotic scale-up using data-driven diagnostics and alert systems.
Engineer state-of-the-art robotic software to increase performance of machine learning.
Design efficient compilation architectures to decouple hardware program inter-dependencies.
Structure docker compose configuration to balance robotic hardware sharing.